ATLAS observes scattering of vector bosons

The exploration of W- and Z-boson interactions at the energy frontier probes the heart of the Brout–Englert–Higgs mechanism. The cross-section of longitudinal weak-boson scattering would diverge, resulting in meaningless values, were it not for the exact cancellation due to Higgs-boson contributions. The key processes for this exploration are the scattering between W and Z bosons emitted by quarks in proton–proton collisions, which are among the rarest processes of the Standard Model (SM) and that have remained inaccessible until very recently.

At the 2018 International Conference on High Energy Physics (ICHEP), held in Seoul on 4–11 July, ATLAS reported the observation of the W^±W^±jj final state, and, for the first time, the observation of the W^±Zjj final state produced by pure electroweak processes, among which vector-boson scattering (VBS) is dominant. Observation of the electroweak production of W^±W^±jj was reported by the CMS collaboration in 2017.

ATLAS data corresponding to an integrated luminosity of 36 fb^–1 collected in 2015 and 2016 at a centre-of-mass energy of 13 TeV were used. The two final states were searched for using W- and Z-boson decays to leptons (electrons or muons), featuring the typical signature of a centrally produced diboson system accompanied by two forward jets that are well separated in rapidity. The large invariant mass (m_jj) of the two jets was used to isolate signal events from the overwhelming background arising from strong interactions. Further selection requirements, utilising additional features in the two channels, were necessary to suppress this background.

In the WWjj channel, the strong-interaction contribution to the production can be greatly reduced by selecting events with the same W-boson charge. Remaining backgrounds arise from processes in which leptons are misidentified or the charge of the lepton is incorrectly measured. The analysis therefore focused on the reduction and control of these backgrounds that are estimated from data. Additional background from incompletely reconstructed WZ events was estimated from simulations. The final m_jj distribution of selected events is shown in the left-hand figure, with the signal accumulating at large m_jj values. The analysis led to a significance of 6.9σ, qualifying for an observation.

Most of the background in the WZjj channel arises from strong-interaction processes contributing to the same final state. Kinematic variables that show distinct differences between electroweak and strong production were exploited to isolate the signal using a multivariate discriminant from a boosted decision tree (figure, right). The analysis leads to an observed significance of 5.6σ.

These observations open up a new era of exploration of a yet largely unknown part of the SM: the quartic couplings of weak bosons. The larger amounts of data collected during LHC Run 2 and future runs will allow for a detailed characterisation of VBS interactions using differential cross-section measurements. Such measurements combined with refined theory modelling provide sensitive tests of the electroweak sector of the SM, and may reveal signs of new physics.